TOMTOM'S Frequency Response ?

[Christiaan_Nel]

Whatsup forum!

I really hope you guys can help me with this one.

I'm a sound engineering student and we have received a project to do a GPS voice over.
Now I have recorded the all commands and I would just like to know what the frequency response of a TomTom Gps is, for Compression and Equalization purposes.

Any help would be greatly appreciated

 

[dhn]

Welcome to TTF.

One of our supermods, Andy_P, happens to be a sound engineer!!!

So, check back into this thread as I'm sure he'll see (and respond) to the post when he logs in.

 

[canderson]

Absolutely ZERO idea. Take a scope to your recorded results and look at the 'crispy' sounds like a "T" and "S" in the command "turn around when possible". A sibilant like an "S" seems the most likely place to find the highest frequencies.

Did you record these from the speaker, or from the audio output of one of the older units, or a Bluetooth audio connection, or ???

 

[Andy_P]

Hmm... interesting question. I presume you mean the basic frequency response of the digital audio system and not the overall response including the analogue amp and loudspeaker?

Simple answer: I don't know any of the hardware specs.

But what I have just done is take one of the "standard" TomTom voices, (UK voice Jane - data05.chk) and used "VifTool" to split it into the 58 individual sound files ("turn right" etc.)

Each of those is a 22050Hz 32-bit mono OggVorbis file, so I suspect that 22050Hz is going to be the native frequency response of the device electronics.

Next I put one of the individual files (I used the phrase "Six hundred" so it has plenty of sibilant HF energy) and put it into Adobe Audition's Spectral View display....

As you can see, there is plenty of useful audio upto around 11kHz (scale on right) as you'd expect from the 22050Hz sample frequency.

Hope that helps... If you've any other questions I'll do my best to answer!

I can also help with the tools needed to actually compile a working TomTom voice

 

[Wyvern46]

I'm very impressed - you need your Genesis strapline back...

 

[Andy_P]

I need a fricking sound job....

 

[Wyvern46]

Economy seems to be picking up..I'll keep my fingers crossed for you.

 

[dhn]

http://www.tomtom.jobs/apply.aspx

Select London and all types of jobs.

 

[Christiaan_Nel]

Thanks so much for the quick reply!
really helpful info, but I should have been more specific... I'm looking for the frequency response of the speakers that the TomTom uses, so I can see what the lowest frequency is that it can produce. So I can EQ it out. If it can't reproduce those frequencies there's no use having them there, right?
So any guesses as to where this frequency is would b e great. I've been looking around and I can't find the types of speakers TomTom uses. So Ive just been cutting everything below 500Hz, as these are very low frequencies and I doubt that such a small speaker can reproduce them.

But thanks a million guys really appreciate the help.

 

[canderson]

Ah, that's why I was asking what sort of output method you were planning to use and to measure. However, as Andy has discovered from at least the ogg style files (no idea if the Loquendo voices follow suit), 11KHz is the top end of what he saw. Nothing below 500Hz would be necessary to make the devices intelligible, and as small as they are (and they don't feel/appear to be very compliant anyway), there can't be much happening down there.

The little speakers are certainly far more likely to be able to reproduce 11KHz than they are anything at the bottom end, but you're in the same boat as the rest of us. All we've ever been able to do (sometimes) is to provide a user with a useful substitution with another speaker of the right impedance and size when one goes kaput. We've got no specs on the things at all as far as frequency response goes.

 

[Andy_P]

Personally, I'd have gone a bit lower than 500 Hz, maybe somewhere around 250? I disagree that nothing useful is happening at low crews - some of the older 'fatter' TT models have better speakers and larger cases, so sound MUCH nicer.
Don't forget many sat-nav systems have line outputs or can transmit the navigation voice to the car's audio system via Bluetooth or FM, so higher qualities CAN be utilised


As a comparison of current models, the Go1000/2000 range sound very good considering how thin they are, but the larger START60 sounds truely awful - really thin and tinny.

I don't have the equipment here to do proper acoustic measurements, but to give a rough idea, I'll record a frequency sweep and play it back over a TomTom Go720, then report back what I hear.

If you have the equipment at college maybe you could do the same and measure it properly? (you'd need to use an old model TomTom or Garmin with a built-in mp3 player.

 

[Andy_P]

Well I tried some step tones on my old Go720 (as it has the mp3 player)

Not a very scientific test but for the first tones of 20, 30, 40, 50, 60, 80 Hz I could definitely hear SOMETHING from theTT loudspeaker for them all (presumably harmonics, but you could tell what each of the tones were.

The next tone was 100Hz and VERY INTERESTINGLY this was completely silent from the TomTom.

The following tones were 250,500 and 1000Hz and they were all played loud and clear.

So it seems there is some savage filtering at 100Hz. Common sense would say it would be a high pass filter so nothing below 100Hz would be getting through either, but if that's the case then the harmonics from the lower frequencies below 100Hz must be getting generated before the filtering, otherwise I would have heard silence for ALL the frequencies at or below 100Hz.

 

[canderson]

Assume you were using a sine wave for all of this, and not much chance of hearing the harmonics from the lower freqs you were throwing at it? As you say, strange about the 100Hz.

As to 'not much happening', that wasn't meant to say that nothing might be produced down there, just that it might not add any to the intelligibility of the sounds the user receives. Some of the 'tinnier' sounding units don't seem to have anything down there at all.

One thing you wisely pointed out .. there's not just one speaker being used across the range, and the differences in case designs create very different sounds (particularly at the lower end) and resonances.