Jump to content

SpecLab Waterfall Scene Capture Tutorial


Recommended Posts

When analysing a full film as a wav file, once the film has finished is it possible to scroll down the waterfall to find the best bits?

Once you open the wav file in Spectrum Lab you will get a file analysis settings window. Change the speed to "fast, occasional waterfall updates." Once the waterfall starts, right click the screen and turn off the Amplitude Bar. Now the entire movie will go through Spectrum Lab in about 2 minutes. You can use the pause button on the left to stop and get playback areas for further analysis. The top button on the left is the progress indicator and it rotates through several pieces of information. It will show exactly how far you are into the movie. Once you write down a timestamp, you can can then click the Continue button on the left to keep analyzing. I have found this method to be an easy way to quickly find low bass content. 


I then use JRiver to create the short clips by changing the playback range and using Convert Format. Sometimes I turn off any DSP when converting so I can get the original 6 or 8 channels for viewing in Audacity to see clipping or how they were mixed. This way I don't have to convert the entire movie into a multi-channel wav. 

Link to comment
Share on other sites

The tutorial above is for scenecaps with a signal chain (through a BD player or AVR subwoofer out).  For the wav files made by JRMC, I use the following settings (attached).

Your peak hold time isn't long enough for movies like LOTR and The Hobbit. You have it set to 9999 which is 166.65 minutes.

Link to comment
Share on other sites

What do you use to convert the audio to individual wavs? I'm using eac3to but when it detects clipping it adds a negative gain. Is that a problem? Its only 1 or 2db...

You'll find that a lot of MKV rips will give you this problem in eac3to.  In fact, you'll get lots of errors/notifications in the eac3to logs when trying to extract audio from MKVs.  For this reason, it is only recommended to use the full audio off the retail discs that you own, especially when posting content here.


How do I add 30 seconds of silence before and after the wav file? Once I have this nailed I can start posting content as I have a ton of films on my server :).


Basically just use whatever WAV editing software to add the blank space before and after the track.  I think any of the mixing software mentioned thus far has the capacity to do it, and Google will provide tons of tutorials to get you started.


As above, this is always welcome, but please refrain from posting analysis of "rips."  I recommend only focusing your analysis on the full, uncompressed audio from discs you have in your possession.  :)

Link to comment
Share on other sites

  • 1 month later...

First, let me say I appreciate all the hard work put into this tutorial and all the measurements.  As more subs are tested outdoors here and more people are taking in room measurements, I really think it is time more people start doing mic'd speclab captures to verify the performance of their sub systems with real world content.  I started a thread about this at AVS awhile ago, but nobody seemed that interested.  I would really like to get this working and put together a similar tutorial so more people can take captures in a way they can be directly compared.


At this point, I think I have figured out a way to add my mic compensation into speclab using a filter, but I am unsure how the levels should be calibrated so the results can be compared to the digital captures in this forum and so they can be compared between systems with different equipment in different rooms.  Can I use a similar process as the one in this thread?  Possibly something like this?


Calibrating for Reference Level:


Play Soho54's Audio Test DVD.

1. Go to '5.1 DD Tests'

2. Go to 'Manual Tones'

3. Go to 'LFE'

I assume you have SpecLab up and running, and have set up your AVR or BluRay to send a bass-managed signal to your soundcard.  You will now use the Spectrum Graph Area at the top of the screen.

4. Play the 60Hz tone.  It is encoded at -20dBFS, which means that through a properly calibrated AVR, it will have a level of 95dB on an SPL meter.  This may be an issue because many people will not want to play a tone at 95db.  May need to find a lower level tone or set the MV to -10 or -20.  Also, if the FR is not flat at the measurement position, 60hz may not be the best frequency to play.  Each person may need to choose a frequency that is at the "flat" part of the FR.


5. You should see something like the screen attached below (60Hz Calibration) for the duration of the tone.

You will now adjust the volume on your AVR so that the 60Hz tone will be right around the -30dB line (see attached pic).

The adjustments should be made on the mic input level and speclab offset value, not the AVR.


6. Go back to the 5.1 DD Tests page.

7. Go to 'Extras' 

8. Go to '-0.5dB LFE'

9. Play the 20Hz tone.

The tone should just touch the -10dB line.  See the attached pic.  In the attached pic, you can see three things:   

   A. The tone is at -10dB on the upper scale.

   B. There was a 'pop'  a few seconds after the tone began, that is my WiFi card....I turn it off when I am running PvA.

   C. There is some 3rd HD and 5th HD in the trace.  I am not clipping my soundcard input, nor my receiver out (verified independently), but the soft clip is at the bluray player, or encoded on the disc, I'm not sure which.  It is still a good way to calibrate.

Why -10dB?  Because the worst case scenario for a 5.1 track is if every single channel has 0dBFS LF material encoded in it, you end up getting 125dB.  This ensures your soundcard will just start to clip if the worst case scenario actually happens.

Not sure if these steps are absolutely necessary.  Again, most people will probably not want to play a -.5dbfs 20hz tone through their system.  Instead, could we play the tone but drop the MV to around -30?  Then speclab should show the level at around -40db. (depending on the system's FR)


You are now almost ready to capture scenes.  WRITE DOWN your Soundcard Input and AVR volume settings, so that you can have apples/apples comparisons.  Always set them the same, and re-calibrate every now and then with the Soho DVD.

At this point, we would write down the mic input levels and speclab offset as the AVR MV should be at 0 for testing.



Two last questions.  First, for people that don't listen at reference, could the speclab offset be changed by the same amount as the MV volume is set under reference?  This way the color palette can still be directly compared to the digital captures.  For example, the speclab offset is set to 25 for a reference level capture and the person normally listens at -15.  If  the offset is increased to 40, will that accurately compensate?  Lastly, is there a way to clear the speclab screen and buffers without restarting the program?  


I appreciate any input.

Link to comment
Share on other sites

  • 1 month later...

I have tried getting my speclab to work for movies but have never got it up and running ever. Its been a year now. So I have a disc made from Soho's link. I was wanting to know can I use the BR player in my computer? If so How do I set that up?


If not I have a PS3 going to my Denon AVR. I could send the LFE out to my line input on my computer if this is easier than using my built in BR player. Either way any help would be great. I am hoping I cant finally post some screen shots of ULF bass in movies.



Link to comment
Share on other sites

  • 1 month later...
  • 4 months later...


This topic is now archived and is closed to further replies.

  • Create New...