Sokel
Grand Contributor
- Joined
- Sep 8, 2021
- Messages
- 11,137
- Likes
- 12,086
You can do that with the mouse wheel, just point it out of the chart to the axis you want.
-
Welcome to ASR. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
Beta Test: Multitone Loopback Analyzer software
- Thread starter pkane
- Start date
morillon
Major Contributor
- Joined
- Apr 19, 2022
- Messages
- 2,151
- Likes
- 535
Thank you !!!!
Super!!
I didn't say anything...and I said it badly too!
hehe !!!!
;-)
(I look like a complete idiot! But it's also because I rarely use the mouse that came with my laptop.If I can do without it in rew... I just need to use it properly with mt ;-) )
Super!!
I didn't say anything...and I said it badly too!
hehe !!!!
;-)
(I look like a complete idiot! But it's also because I rarely use the mouse that came with my laptop.If I can do without it in rew... I just need to use it properly with mt ;-) )
Sokel
Grand Contributor
- Joined
- Sep 8, 2021
- Messages
- 11,137
- Likes
- 12,086
Practice makes perfect!
morillon
Major Contributor
- Joined
- Apr 19, 2022
- Messages
- 2,151
- Likes
- 535
If the differentiated L and R " cal" correspond to a somewhat specific use (here, a headphone measurement bench...),
a fairly common smoothing option , into the "fft gain ave freq" bar would be really welcome ( in mode "spectrum" or "frequency reponse " only (?) )...
and darn it, I think I'm not completely "idiot" for this "need"... ;-)))
a fairly common smoothing option , into the "fft gain ave freq" bar would be really welcome ( in mode "spectrum" or "frequency reponse " only (?) )...
and darn it, I think I'm not completely "idiot" for this "need"... ;-)))
Last edited:
Hi Paul,
I wonder what is the maximal FFT length one can use with DSD512 GEN? Is it computer power limitation only or there is software limitation?
I am trying to run AP32, DSD512, 4M.
Meanwhile 2 and half hours...
It is strange that the task manager doesn't show something special in activity.
Of course, before running the 4M I ran simple 128k to see everything is working properly.
Thanks,
Guy
I wonder what is the maximal FFT length one can use with DSD512 GEN? Is it computer power limitation only or there is software limitation?
I am trying to run AP32, DSD512, 4M.
Meanwhile 2 and half hours...
It is strange that the task manager doesn't show something special in activity.
Of course, before running the 4M I ran simple 128k to see everything is working properly.
Thanks,
Guy
OP
- Thread Starter
- #4,746
Hi Guy,
There's no built in upper limit in Multitone, in fact, almost all settings can be entered by hand and set to any value you choose. Of course, some may require a lot more memory and processing power. How many averages and at what overlap are you trying to generate? This determines the length of the test signal, which is then converted to DSD.
Hi Paul,
I tried DSD512, 4M, 16 averages.
I knew it might take forever...
I ran about 5 hours and then I stopped it.
The computer and program was fine, after stopping it I ran short test and it passed.
Any rough estimation of how much time one average of DSD 512 4M could take?
I ran it on Ryzen 5 pro, 6650H with 16GB ram.
I thought that the long time is mainly the first average.
I don't know how MTA making DSD, but for example with Audirvana software I can convert PCM 48k to DSD 512 with 7 order modulator on the fly.
I don't know if this method gives 'real' DSD. Of course DSD, but just a way to encode the PCM without the benefits of DSD? This method gives very good measurements with my DAC, but of course I need to calculate myself, the MTA or REW don't calculate numbers from external source.
Guy
I tried DSD512, 4M, 16 averages.
I knew it might take forever...
I ran about 5 hours and then I stopped it.
The computer and program was fine, after stopping it I ran short test and it passed.
Any rough estimation of how much time one average of DSD 512 4M could take?
I ran it on Ryzen 5 pro, 6650H with 16GB ram.
I thought that the long time is mainly the first average.
I don't know how MTA making DSD, but for example with Audirvana software I can convert PCM 48k to DSD 512 with 7 order modulator on the fly.
I don't know if this method gives 'real' DSD. Of course DSD, but just a way to encode the PCM without the benefits of DSD? This method gives very good measurements with my DAC, but of course I need to calculate myself, the MTA or REW don't calculate numbers from external source.
Guy
OP
- Thread Starter
- #4,748
Not sure why you need 4M FFT for DSD generation, Guy. It's used as part of the upsampling process to DSD512 in MTA, and there's little reason for such a large/sharp filter. MTA generates the full signal in memory before it's played back, which is why I suspect that you are just running out of memory, rather than getting bogged down with calculations.
Yes, I don't need, just wanted to test limitation.
I did not see that the RAM memory is fully used (in task manager).
When I want to reduce (or measure) skirts I am using 4M. This is a suggestion I got from some member and indeed it is correct specially if measurement is not synchronic (ADC-DAC).
OP
- Thread Starter
- #4,751
By the way, you can measure external signals with MTA (and I believe, with REW, but I could be wrong). With MTA, if it's a single or multiple tone test, you'll just need to set output to EXTERNAL and set the threshold for detecting fundamental in settings:
OP
- Thread Starter
- #4,754
The signal is generated as PCM at some specified multiple of the base DSD rate, and then upsampled and converted to 1 bit PDM. DSD Gen Multiplier setting under Test Signal determines the native PCM rate that MTA uses to generate the test before it's upsampled to the desired DSD rate. Multiplier of 1 means that PCM will be generated at 48k, 4x means 192k, etc.
OP
- Thread Starter
- #4,755
The limits depend a lot on what you're trying to do, what settings you chose, and the hardware memory and processor (especially double-floating point processor) that is used. On laptops and mobile CPU processors, double floating point may not be natively supported. All calculations in MTA will be come significantly slower, since internally everything is represented as double.
OP
- Thread Starter
- #4,756
Just tried on a 16GB Windows VM, 4M FFT, AP 32 test signal, and indeed, it runs out of memory at 16 averages. You can test it by selecting File->Save Test Signal... and choosing different lengths. The following gave me an out of memory error:
But the following generated DSD512 and saved it to a file with no issues, took less than a minute:
But the following generated DSD512 and saved it to a file with no issues, took less than a minute:
Last, so, is there a difference between playing DSD from Audirvana (upsampled from PCM) than doing it directly by MTA?
I understand that with the threshold function the MTA will be able to calculate THD(+N) for external source (one tone), but what about IMD of AP32 with external source? Is it possibly for the MTA to compute?
I understand that with the threshold function the MTA will be able to calculate THD(+N) for external source (one tone), but what about IMD of AP32 with external source? Is it possibly for the MTA to compute?
OP
- Thread Starter
- #4,759
I don't know what Audirvana does, so can't tell you the difference. MTA detects any number of fundamental tones in the recorded signal, as long as the threshold is set appropriately. The idea of threshold is to set it high enough to allow all the fundamental tones through but to reject any IMD or other spurious signals. When set correctly, all 32 tones will be detected and IMD will be measured just as if MTA played the signal.
Similar threads
