Shredhead Posted June 6, 2015 Report Share Posted June 6, 2015 Soundtracks have transients under 40 ms all over the place, and there's no problem measuring these. I think the whole thing that mojave (correct me if I am wrong mo) is trying to achieve is to find a program or a setting in SL that finds the accurate SPL of transients in the content. He mentioned something about a plugin called spectrafoo or something like that but I am not familiar with it. Link to comment Share on other sites More sharing options...
maxmercy Posted June 6, 2015 Report Share Posted June 6, 2015 I think the whole thing that mojave (correct me if I am wrong mo) is trying to achieve is to find a program or a setting in SL that finds the accurate SPL of transients in the content. He mentioned something about a plugin called spectrafoo or something like that but I am not familiar with it. No such thing in SL I am aware of. So much depends on calibration, and what 'reference' is. As for 40ms transients, that is large enough timeframe to hold a full cycle of a 25Hz waveform, and many sharp explosions (esp the 'crack' and 'boom' portion, but not the 'shake/shudder/rumble') have their highest SPL within that intial 40ms window. Danley Fireworks comes to mind. But without the following shudder/rumble that can last over 6x-10x that 40ms time, they will lose some impact. JSS Link to comment Share on other sites More sharing options...
SME Posted June 8, 2015 Report Share Posted June 8, 2015 No such thing in SL I am aware of. So much depends on calibration, and what 'reference' is. As for 40ms transients, that is large enough timeframe to hold a full cycle of a 25Hz waveform, and many sharp explosions (esp the 'crack' and 'boom' portion, but not the 'shake/shudder/rumble') have their highest SPL within that intial 40ms window. Danley Fireworks comes to mind. But without the following shudder/rumble that can last over 6x-10x that 40ms time, they will lose some impact. JSS If the transient contains high frequency as well as low frequency content, then the transient can occur in a much shorter span of time and still have content to low frequencies. Effects with high SPL and short time slam very impressively if reproduced well, but because it takes a lot of SPL to make it happen, most movies heavily compress the bass sound effects, effectively elongating the transients, to provide loudness in place of SPL. For the same SPL the movie sounds are perceived to be a lot louder than their real-life counterparts. Nevertheless, there's no mistaking the real sound if one has a chance to compare the two. On the subject of fireworks, I really hate the Disney fireworks trailer. It sounds fake and pointlessly loud. The fireworks in Oz are pretty good as far as movie sounds go. I think they are a good compromise. Then there's the Danley fireworks recording, which impresses me a lot. Even though it doesn't seem especially loud when played back with peaks in the high one-teens, the slam is truly first rate. When I have upgraded sub capability, I'll have to give it a shot with a bit more SPL and hope I don't break anything. I wish movies did sound more like the Danley fireworks, even though they would have to compromise on loudness in order to fit it into the 125 dB (or 123 dB, going by evidence that 83 dB is the true reference for -20 dBFS RMS pink noise). Perhaps some day the other 99.99% of the population (or at least theatrical installations) will have bass systems that can peak at 130 dB and reproduce some of these powerful yet highly transient sounds realistically. Link to comment Share on other sites More sharing options...
maxmercy Posted June 8, 2015 Report Share Posted June 8, 2015 With a few mouse-clicks and Audacity, you can easily create a 7.1 Danley Fireworks track. Then you have 128dB peaks (if you dare to playback at reference). Unfortunately, both tracks available are clipped on the loudest booms, but only the loudest ones. Tom was going to look at it a while back, but I never heard back. Danley tracks available here: http://www.diysoundgroup.com/forum/index.php?topic=389.0 Also a smaller, terrific forum with much more signal than noise. JSS Link to comment Share on other sites More sharing options...
SME Posted June 8, 2015 Report Share Posted June 8, 2015 With a few mouse-clicks and Audacity, you can easily create a 7.1 Danley Fireworks track. Then you have 128dB peaks (if you dare to playback at reference). Unfortunately, both tracks available are clipped on the loudest booms, but only the loudest ones. Tom was going to look at it a while back, but I never heard back. Danley tracks available here: http://www.diysoundgroup.com/forum/index.php?topic=389.0 Also a smaller, terrific forum with much more signal than noise. JSS Or one can just play back the stereo recordings at a level well above reference. That's better anyway because the recordings have strong directional cues from the recording technique that only render correctly in stereo anyway. The clipping is tragic but is indicative of how hard it is to record real-life high SPL events. I think the only solution is to use multiple mikes with drastically differing sensitivity and a mix-down to widen the dynamic range. Then correct playback becomes even more difficult. On the subject of real life recordings, I see no one has attempted to obtain Bob Katz's Space Shuttle launch recording. Perhaps that's because it's not readily downloadable. Katz asks that you create an account on his Digital Domain site to download it. I made a 5.1 channel .FLAC version of it from the stereo .WAV pair he provided, and he put it on his site. To my ears at least, it does not have any clipping at all, and is very impressive. I think those of you with real ULF capability as opposed to the mere 16 Hz capability I have will be much more impressed. I'm probably compressing pretty heavily when playing it back. I know that when I get around to building my new subs, it will be one the first tracks I go to in order to assess what playback levels I can reach. It is 4 channels played at +7 dB above 83 dB RMS reference after peak normalization, so its absolute peak levels should approach 125 dB or so. That's very close to 5.1 dBFS (at 126 dB, using 83 dB RMS for calibration). According to Katz's measurements, the PvA peaks at 25 Hz and stays hot all the way to DC. WOTW, step aside. Bosso Dave, you should download and play the Space Shuttle at your GTG. Be sure to sign up on the forum ahead of time so you get approval in time. Edit: add link to spec/PvA Link to comment Share on other sites More sharing options...
maxmercy Posted June 8, 2015 Report Share Posted June 8, 2015 The space shuttle recording does not have an LFE channel encoded, IIRC. I have upmixed the Danley FW track to 7.1 with great success, no loss of directionality by my ears, and a firm front soundstage with only a little envelopment to sides, very little to rear. JSS Link to comment Share on other sites More sharing options...
SME Posted June 9, 2015 Report Share Posted June 9, 2015 The space shuttle recording does not have an LFE channel encoded, IIRC. I have upmixed the Danley FW track to 7.1 with great success, no loss of directionality by my ears, and a firm front soundstage with only a little envelopment to sides, very little to rear. JSS No, the Space Shuttle track does not have an LFE channel, but it is 4 channel and is intended to be played at +7 above reference. That allows a peak level only 1 dB below the 5.1 dBFS peak level. The track is peak normalized, and because it was recorded with 4 mics placed relatively close together and is very bass heavy, the peak level is probably very close to full-scale level with all channels summed. In contrast a typical 5.1 track typically doesn't typically have much bass in the surrounds. That trend is changing with Atmos mixes where the target system, i.e. the theater, gets bass management for surrounds, but in any case, I do believe that this space shuttle recording is competitive with WOTW and such as far as the demands it places on the system to properly reproduce. If I'm reading and interpreting his spectrogram right, it wants 115 dB at 8 Hz or 118 dB at 10 Hz! I'm curious as to why you would want to make 7.1 channel versions of the 2 channel Danley recordings tracks. Does your system restrict your ability to chose a higher than reference playback level? Or are you trying to enhance the presentation, similar to using PL2x to upmix from 2 channel to surround? How did you create it? Did you just copy left to FC, FL, SL, and BL and right into FC, FR, SR, and BR, or did you use some kind of up-mixing plugin? FWIW, I used to listen to most of my music with PL2x, but since ditching Audyssey MultEQ XT for my own DRC, I've been listening to 2 channel content in stereo. I can run Audyssey DynamicEQ in L/R "bypass" mode and still enjoy it's benefit with music. With movies, I usually aim for "reference level" in a subjective since and don't need any bass boost. My FL and FR also have the best cleanest in-room response out of all my speakers. After adding my room treatments, 2 channel imaging improved dramatically, significantly more than for 5.1 where the additional speakers helped compensate for the acoustic problems. In fact, when sitting in the sweet spot, the positional imaging is very similar to what I heard with PL2x. The surround has a bit less emphasized, but that's mostly because I don't have Dynamic EQ overboosting my surrounds like it did before. The surround is also quite a bit more diffuse, but it's supposed to be like that. I still get some imaging all the way around my head when playing in 2 channels. Getting back to the Danley recordings, I have noticed that playing the with PL2x really messed with the imaging. Particularly, the parade recordings just sounded odd. Like, even though things would image strongly the images would cycle across the sound stage repeatedly. I don't think I bothered to try them in 2 channels to compare until I got my room treated, and I do think they sound much better in stereo than with PL2x. Link to comment Share on other sites More sharing options...
maxmercy Posted June 10, 2015 Report Share Posted June 10, 2015 Straight up mix, copy L/R to surrounds and mix for center. Then apply BEQ, and add a touch of mid bass until it sounds like a real fireworks presentation. Torture test for subs and LCRS... JSS Link to comment Share on other sites More sharing options...
SME Posted June 10, 2015 Report Share Posted June 10, 2015 Straight up mix, copy L/R to surrounds and mix for center. Then apply BEQ, and add a touch of mid bass until it sounds like a real fireworks presentation. Torture test for subs and LCRS... JSS Why BEQ for the Danley fireworks? Is it filtered? By the way, there's actually another Danley fireworks recording out there that's a bit older. I'm not sure how BEQ would make it more like real fireworks. I think it sounds pretty close to real with a high enough playback level. Albeit, they still a good distance away at the levels I can play it at. I analyzed the spectrum of just the part of one of the rockets launching and observed heavy energy between 5 and 40 Hz that peaks around 16 Hz. Subjectively I think the bass on those sounds was pretty realistic and impressive, and I'm rolled-off below 16 Hz. As such, I can't imagine needing more ULF. Link to comment Share on other sites More sharing options...
maxmercy Posted June 10, 2015 Report Share Posted June 10, 2015 It didn't 'need' it. I just wanted to play around. The old FW file has a better finale. The newer one is pretty good, though. JSS Link to comment Share on other sites More sharing options...
mojave Posted June 10, 2015 Report Share Posted June 10, 2015 The point here, though, is that the ULF, that is <20 Hz content, will not have a disparity of 15dB from 40 Hz or 100 Hz, or whatever it is you've claimed. You really need to just retract that stuff and face reality. I claimed 10 dB. Link to comment Share on other sites More sharing options...
mojave Posted June 10, 2015 Report Share Posted June 10, 2015 Take the HULK scene you posted... What is funny about this is that you do a +20 dBFS offset in Spectrum Lab. There is no possible way that signal can be so close to 0 dB and still leave room for the rest of the bandwidth. And you are incorrect about "They're saying that this . . . is really this" since no one has said that. Link to comment Share on other sites More sharing options...
mojave Posted June 10, 2015 Report Share Posted June 10, 2015 I tested a clip of War of the Worlds using REW to measure SPL: I used from 0:20:24 to 00:27:50 in War of the Worlds for a test clip. I used an 80 hz crossover with a 24/db per octave slope for all channels (to match the THX filter and speaker sum). I routed the subwoofer channel to REW's input digitally using the MOTU mixer on my MOTU 1248. I used REW with ASIO input to receive the subwoofer signal. I set levels in REW per Blue Sky's http://abluesky.com/support/blue-sky...on-test-files/ test tones and methodology to make it easy to reproduce. The REW SPL Meter was set to C weighting and slow. SPL by Frequency - I used the REW RTA in RTA 1/48 octave mode. With an FFT length of 131072, each FFT bin is only .18 Hz wide. I also used flat top windowing to provide accuracy to .01 dB. From the REW manual - "If the spectral peak amplitudes must be accurately measured use the Flat Top window, this will provide amplitude accuracy of 0.01 dB regardless of where the tone being measured falls relative to the bins of the FFT. " Averaging was set to None to accurately track the peaks. SPL by Time - I used the REW SPL logger with the SPL Meter set to Z weighting and fast. From the manual, it looks like this updates every 125ms. I ran the clip through about 10 times and came up with the same peak numbers every time even though the start time differed slightly each time from when I started playback of the clip. Below are two screen shots of what was on my screen during playback of the clip. Next are SPL by Time and SPL by Frequency. The peak level during playback was 119.1 dB. The highest SPL level at any given frequency was at 30 Hz with an SPL of 104.9. The peak SPL is 119.1 because it is a combination of the SPL of all the bass frequencies combined at once. The SPL levels get higher below 80Hz as the redirected bass is added to the LFE channel. The dynamic range during this clip is 80.1 dB (119.1-39) The data starts at 15 Hz. From 15 Hz and up, the peak level at any given frequency does not exceed the measured capability of the SVS PB13 per data-bass.com without even factoring in room gain. The peak SPL of 119.1 dB also does not exceed the measured capability of the SVS PB13. We don't know from this data if any transients were missed, what distortion is present, or what frequencies are being played at any given time. That data is all obtained using Spectrum Lab. Link to comment Share on other sites More sharing options...
Bossobass Dave Posted June 12, 2015 Author Report Share Posted June 12, 2015 I claimed 10 dB. No, you didn't. Maybe read your own posts? What is funny about this is that you do a +20 dBFS offset in Spectrum Lab. There is no possible way that signal can be so close to 0 dB and still leave room for the rest of the bandwidth. When you mic a subwoofer, the levels can be wherever you've calibrated your sub to be. And you are incorrect about "They're saying that this . . . is really this" since no one has said that. Sure they did. Cherry pick a scene from WOTW and do what with it? Prove a ported sub can play some of it back? Did you remember to tilt the graphs? That's really useful information. Good work. Link to comment Share on other sites More sharing options...
mojave Posted June 12, 2015 Report Share Posted June 12, 2015 When you mic a subwoofer, the levels can be wherever you've calibrated your sub to be. I'm using data directly from the disc which shows the accurate levels as they are encoded. Cherry pick a scene from WOTW and do what with it? Prove a ported sub can play some of it back? I'm not sure what you mean by "cherry pick." I used the "Emerge of the Pods" scene that is on the SuperLeo demo disc and a few other demo discs. I own the War of the Worlds Blu-ray and provided the exact time stamps and did all my testing from the Blu-ray rather than a demo disc to make sure the levels were correct. The duration of the clip is 7 minutes 26 seconds. It is longer than any of the clips I have seen used here. I thought the clip was universally accepted as a good bass clip and of long enough duration to not be considered "cherry picked." None of the graphs I ever looked at supported claiming anything more than a 10 dB difference so I don't know why I would have claimed it. The only quote I can find says, "Here is The Dark Knight Rises tilted. You can see that it should be down in-room by almost 10 dB at 20 Hz vs 40 Hz." Do you think REW is inaccurate in its SPL readout? It showed peaks of 119.1 dB which seems accurate to me for a 5.1 channel track with rerouted bass. The bottom octave of REW's RTA is from 15.625 Hz to 31.25 Hz. It divides this into .3255 Hz bins for the 1/48th Octave RTA. The FFT length used by the RTA is 131072. This corresponds with .18 Hz bins that are actually being measured. I used Flat Top windowing, which the REW manual says is the only way to accurately measure spectral peaks. The other windowing options can be off by as much as 3.92 dB. I did not use the Adjust RTA levels since that is only for pink noise or other test signals. From the REW manual: The RTA plot shows the energy within each octave fraction bandwidth. As the RTA resolution increases, from 1 octave through to 1/48 octave, the octave fraction bandwidths decrease and, for broadband test signals such as pink noise, the energy in each octave fraction decreases correspondingly. Whilst the RTA is correctly showing the actual level within each octave fraction, this variation of trace level with RTA resolution can be awkward when using the RTA with a pink PN noise signal to adjust speaker positions or equaliser settings. The Adjust RTA Levels option offsets the levels shown on the RTA plot to compensate for both the bandwidth variation as resolution is changed and the difference between a sweep measurement at a given sweep level and a pink PN RTA measurement at the same level, allowing direct comparison between RTA and sweep plots. Whilst the levels shown are not the true SPL in each octave fraction, they are more convenient to work with. N.B. This option should only be used with broadband test signals, pink noise or pink PN. If you have a better methodology or improvements to accurately show SPL down to .3255 Hz bins, then I'm open to suggestions. This is the first time I've seen a movie measured through REW with a fully digital signal chain. I think it is pretty cool that one can run an entire movie through REW and find from the logger where the peak SPL is located and what the actual peak is from that movie. It also gives a good picture of the dynamics when you see 50+ dB swings over a short interval. Link to comment Share on other sites More sharing options...
mojave Posted June 12, 2015 Report Share Posted June 12, 2015 Here is the opening scene from EOT. One image is using the voltage setting in Spectrum Lab and the other is using the dBFS setting. It illustrates that even the same program with with the same signal can display it in two different ways depending on what you want it to show. Just because the data looks different in each screen shot doesn't mean that one is wrong. An SPL display like an RTA does is another way of looking at the data. Poking fun of an RTA display of the same data is no different than complaining that the voltage measurement by Spectrum Lab is somehow reducing the bandwidth or output. Link to comment Share on other sites More sharing options...
Bossobass Dave Posted June 12, 2015 Author Report Share Posted June 12, 2015 So we agree. The SL caps do not need to be tilted and nothing SVS makes can reproduce WOTW accurately. Link to comment Share on other sites More sharing options...
Shredhead Posted June 12, 2015 Report Share Posted June 12, 2015 I'm not sure what you mean by "cherry pick." The scene you picked is primarily centered at 30Hz and is the easiest action scene in WOTW to reproduce due to it's frequency content. The data starts at 15 Hz. From 15 Hz and up, the peak level at any given frequency does not exceed the measured capability of the SVS PB13 per data-bass.com without even factoring in room gain. REW's RTA doesn't show anything below 15Hz, that does not mean that there is not information below that. Take the airplane crash scene for example: On RTA you cannot see anything below 15Hz: Unfortunately a ported sub will not even come close to reproducing WotW's content. Mo, I don't really get the point you're trying to make anymore. If you want to use REW's RTA to analyze content and not SL, that's just fine with me. Go create a website centered around RTA's for movie content. Link to comment Share on other sites More sharing options...
mojave Posted June 12, 2015 Report Share Posted June 12, 2015 Unfortunately a ported sub will not even come close to reproducing WotW's content. I completely agree. Never said it would. The scene you picked is primarily centered at 30Hz and is the easiest action scene in WOTW to reproduce due to it's frequency content. I said an SVS PB13+ Ultra could playback the content from 14 Hz and up. I also gave a volume level of 118 dB which is below reference level. Feel free to disagree, but don't make it sound like I think any ported sub can play any content. I'm not sure why you guys are making such a big deal about it. From the various screen caps for War of the Worlds posted on data-bass, the one I chose has the most content above 14 Hz making it the most difficult for the subwoofer to handle. If it has a high pass filter, then lower content isn't making it more difficult. Here is the screen cap: REW's RTA doesn't show anything below 15Hz, that does not mean that there is not information below that. Take the airplane crash scene for example: I'm going to ask JM if he can extend it. I think the REW peak SPL meter is still capturing all the frequencies as shown in Bossobass's capture from the GTG. Go create a website centered around RTA's for movie content. Are you starting a website for posting maximum voltage output? Why would every type of analysis need its own website? I spent quite a bit of time coming up with a methodology so anybody could easily use Spectrum Lab to digitally capture the actual content of movies and have helped lots of guys through PM. Why the animosity? Link to comment Share on other sites More sharing options...
Shredhead Posted June 12, 2015 Report Share Posted June 12, 2015 Are you starting a website for posting maximum voltage output? Why would every type of analysis need its own website? I spent quite a bit of time coming up with a methodology so anybody could easily use Spectrum Lab to digitally capture the actual content of movies and have helped lots of guys through PM. Why the animosity? What animosity? My posting maximum V-out includes speclabs with it. I'm not saying because of my scope graph that SL needs some kind of correction rotation because it should match my scope. Here at DB things get rated by FFT. If you think they should be rated by RTA instead then create a site and ratings system based on that. Link to comment Share on other sites More sharing options...
Bossobass Dave Posted June 12, 2015 Author Report Share Posted June 12, 2015 Take the HULK scene you posted... What is funny about this is that you do a +20 dBFS offset in Spectrum Lab. There is no possible way that signal can be so close to 0 dB and still leave room for the rest of the bandwidth. The color scale is relative and can be changed infinitely. When using a microphone to capture the performance by a subwoofer, there is indeed a possible way a signal can be that close to 0dB and still leave room for the rest of the bandwidth. In fact, it routinely exceeds 0dBFS and still has room for the rest of the bandwidth. The color scale is relative and can be changed infinitely (in case you missed it the first time). Changing the offset, shifting the scale, adding colors, etc. will not change the relative aspect. And you are incorrect about "They're saying that this . . . is really this" since no one has said that. This is why some of the current movie graphs at data-bass using actual disc data show a rising tail down below 10 Hz. In the 3-4 octaves below 10 Hz there is an increase of 9-12 dB. That is a Spectrum Lab graph which has a 3 dB/octave rising slope as you go down in frequency... Beginning at 10 Hz, there are 3-1/2 octaves in the bandwidth of interest. That would be 10.5dB plus 12dB = 22.5dB from 1 Hz up. My comment is based on your numbers, which you (that is, someone, not no one) posted (said). Let's see... you've picked at my amplitude bar scale, my offset setting option, accused me of posting in error based on errant reading of your posted data, you've posted doctored graphs, cherry-picked 7 minutes and 26 seconds of a movie to somehow "prove" a particular subwoofer can do the impossible, ignoring the data that clearly shows a much better low tuned ported subwoofer gives nothing below it's tune during playback of WOTW and other movies, accused Shred of hostility... So, awesome... we agree... the SL graphs don't need to be tilted and SVS has nothing that can reproduce WOTW accurately. Can we stop yet? Link to comment Share on other sites More sharing options...
mojave Posted June 12, 2015 Report Share Posted June 12, 2015 Here at DB things get rated by FFT. If you think they should be rated by RTA instead then create a site and ratings system based on that. Here at DB, all the subs were measured using software that shows the transfer function as represented by output in SPL before anything was rated by FFT in the forums. The sorting and ability to compare are all based on SPL. I suggested that since data-bass showed data for subwoofers in SPL, it might make more sense to the AVS crowd to be able to understand a graph that also depicted SPL. Ricci understood it when he said, "The point DD is trying to make is that since the peak and average traces in SL are linear based while most measurements of the response of subs are typically log sine based, that making direct data comparisons of spectral weight/balance of content in movies captured in SL to response charts of subwoofers will be skewed somewhat if the differences in the way the data is presented are not accounted for. Both are accurate but the presentation is different." By the way, aren't both 1. Level and 3. Dynamics measurements based off SPL? Link to comment Share on other sites More sharing options...
Bossobass Dave Posted June 12, 2015 Author Report Share Posted June 12, 2015 Sorry, quoting Josh adds nothing to your argument. We've been pummeling this dead horse for pages now. If Josh meant anything beyond saying that one metric is linear scaled and the other is log scaled, then he is also mistaken. Certainly, the 2 metrics can be and are routinely directly compared. Link to comment Share on other sites More sharing options...
Bossobass Dave Posted June 14, 2015 Author Report Share Posted June 14, 2015 I just saw that JPC is now stalking me through Shreds as a surrogate victim over at AVS. The self proclaimed SpecLab Head Cheese is trashing Adam's thread spraying JPC SpecLab grafiti all over the walls. J says: "Bosso should use full range SpecLabs." Like HE does? Awesome stuff, J, but a little lacking the the low end. Let's zoom in on the SW output... Sorry, still a bit coarse. Let's zoom into the 1st decade, you know, the ULF part? Yeah man, look at that dazzling detail... just like we hear and everything. Super contribution to the Movies With Bass thread. This was like the only one you posted ever, right? Well, what scene is it? Is it from Gone With The Wind or one of your other favorites? But wait, then J, picking at a post of mine in this thread, stealing the copyrighted artwork and posting it elsewhere, pussy that he is, says... "...no one would use those settings..." Really, J? No one? No One. Yep, sounds like the perfect new name. Stay the fuck out of Adam's thread, No One. Find someone new to stalk over there. I'm no longer a posting member at AVS. Start your own thread to wow the crowds, and good luck with that. Link to comment Share on other sites More sharing options...
BeastAudio Posted June 15, 2015 Report Share Posted June 15, 2015 No one can get under your skin quite like that! Whoa! Hehe see what I did there Link to comment Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.