The T.C. Electronic LM5D is an excellent meter if you are looking for ITU-R. B.S. 1770 metering. It gives you a long term number and a short term segment meter, which I like beter than the short term number display on the Media Meter/LM100. It has a peak warning that is user selectable too.
Seems like they both allow logging and variable integration time....
I've used AudioLeak before, its only catch is that it doesn't work in real time -- you have to mix the audio down and then run the audio thru the program, and it spits out a chart and report. The Pro version does do LKFS though.
I quit using LKFS..
Just pre mix your Dia using a VU meter.
-23 LKFS = -3vu
Why have you stopped using LKFS? VU won't give you short-term loudness, integrated loudness or the LRA. Also, I'm not sure -23LKFS would always equal -3VU as the VU meter doesn't use the weighting algorithm therefore it would react to frequencies differently. I'm guessing you got that figure using a 1KHz sine wave?
Also, LUFS/LKFS is designed to be used on the whole programme, not just dialogue.
It also certainly won't give you True Peak metering which more and more broadcasters are now implementing.
+1 for WLM
At work we have the Media Meter but since we got WLM in the update of our Waves bundle I've been using that instead. I really like it. Did a few shows with both it and the Media Meter and the readings are spot on. Plus you get other things like true peak. I really like the read outs of the WLM. Highly recommend it.
A V.I. (volume indicator) which everyone insists on incorrectly calling a "VU" meter, is useful for real time balancing and monitoring of average program levels, but it won't give you LKFS without a lot of visual averaging. So, you need to use some other instrumentation to verify that you are meeting the long-term LKFS specification. On uncompressed speech, if you peak at 0 VU with zero aligned on tone to -20 dBFS, then the long term average speech level will probably be around -27 dBFS. If you are working with bass heavy material and use A-weighting, then you will get erroneous results in comparison with LKFS, since it is basically flat down to 100 Hz whereas the A-weighting is heavily rolled off.