The PSR-E453 is definitely nicer than the PSR-E353. They're very similar in some respects, but different in others. If you can afford it, the PSR-E453 would be a better keyboard to buy than the PSR-E353.
Okay, I put together a "by the numbers" comparison of the PSR-E353 and PSR-E453 from the information in their manuals, in case you're thinking of getting one of them. (Please forgive the periods; they're to make sure things line up nicely.) I didn't include things that are the same, such as the number of keys, reverb types, or chorus types.
................... | PSR-E353 . | PSR-E453
Amplifiers ........ | 2.5 W .... | 6 W
Polyphony (Max.) .. | 32 notes . | 48 notes
Panel Voices ...... | 196 ...... | 237
Drum/SFX Kits ..... | 18 ....... | 24
Arpeggio Voices ... | 20 ....... | 40
XGlite Voices ..... | 339 ...... | 457
Ultra-Wide Stereo . | 3 types .. | No (?)
Panel Sustain ..... | Yes ...... | No
Preset Styles ..... | 158 ...... | 220
External Styles ... | 1 ........ | 10
Preset Patterns ... | No ....... | 25
Pattern Sections .. | N/A ...... | 5
Number of Songs ... | 5 ........ | 10
Number of Tracks .. | 2 ........ | 6 (5 + A)
Data Capacity ..... | 10,000 ... | 19,000 (Approx. notes)
Registration ...... | 9 ........ | 32 (4 x 8)
Music Database .... | 100 ...... | 378
Attack/Release .... | No ....... | Yes (Main and Dual Voices)
Cutoff/Resonance .. | No ....... | Yes (Main and Dual Voices)
Pitch Bend Wheel .. | No ....... | Yes (+/- 1 to +/- 12 semitones)
Assignable Knobs .. | No ....... | 2
DSP Effects ....... | No ....... | 10 types
Scales ............ | No (ET) .. | 5 types (plus custom)
USB Audio I/O ..... | No ....... | Yes
Following are some comments about the preceding information.
Maximum polyphony can be important if you're playing with an auto accompaniment (style) or using a sequencer or MIDI file to play the keyboard, as well as if you're playing with sustained notes. However, the actual polyphony may be less than the maximum depending on the voices selected and whether the Dual option is being used to layer voices together, because some voices use only 1 element and others use 2 elements per note. An element is like a synthesizer's oscillator; it's a tone generator that plays a sound sample for a single note. Thus, the PSR-E353 can play 32 elements at once, which could work out to 32 notes, or 16 notes, or 10 notes, or 8 notes, depending on whether the Main Voice you've selected uses 1 or 2 elements per note, and whether or not the Dual Voice is on and whether it uses 1 or 2 elements per note. In contrast, the PSR-E453 can play 48 elements at once, which could work out to 48 notes, or 24 notes, or 16 notes, or 12 notes. Sustained notes and the auto accompaniment will also use up some of the 32 or 48 elements.
Speaking of sustain, the PSR-E453 doesn't have a "panel sustain" function like the PSR-E353 does, but you can still sustain notes with an attached foot switch or foot pedal, or you can sustain a voice's notes without a pedal by increasing the voice's release time setting.
The stated number of preset voices can be a little bit misleading, because some of them are "Dual voices" that combine two of the other preset voices together, but the PSR-E453 still has a greater number of unique panel voices and XGlite voices than the PSR-E353 does.
Likewise, the number of Arpeggio voices (and Harmony voices, which aren't listed above) aren't important, since the Arpeggio or Harmony option can be used with any voice, and both models have the same number of Harmony and Arpeggio types.
If you like to play with an auto accompaniment (style), the number of preset styles included with the keyboard can be important, as well as the number of external styles you can load (or "register") at the same time.
It's better (i.e., more flexible) to record songs using a DAW, but if you want to use the built-in song sequencer to record then the PSR-E453 lets you record twice as many songs at once and more tracks per song.
The Music Database can be useful for automatically setting up the keyboard for a particular song, because when you choose an entry from the Music Database it selects a specific style number, style variation (Main A or Main B), tempo, Main Voice number and settings, Dual Voice number and settings, and other parameters (Reverb Type, Chorus Type, Harmony Type, etc.).
The Registration memories are also useful for saving and recalling your own setups. The PSR-E453 has 4 buttons for instantly recalling 4 different registrations, and you can select between 8 banks of registrations, giving you 32 registrations in all. The PSR-E353 has 9 registrations in all, and to recall a registration you must press 2 buttons, so it's a little more cumbersome to recall different registrations during a performance.
The PSR-E453 has functions that let you modify the attack/release times and cutoff/resonance settings of the Main and Dual Voices, making it possible to dramatically alter the built-in voices and create "new" voices from them, which can then be saved to the registrations for later recall. You can also make changes to those parameters while performing by using the 2 Live Control knobs. The PSR-E353 doesn't have those features-- although it responds to MIDI messages for changing the attack/release and cutoff/resonance, so you can still modify the voices but only by using an app on an attached computer or iPad/iPhone, or via a MIDI file.
Likewise, the PSR-E453 has a pitch bend wheel, whereas the PSR-E353 doesn't-- although it responds to MIDI pitch bend messages.
The PSR-E453 also has DSP effects, which are new to the PSR-E4xx line. The PSR-E353 doesn't.
The PSR-E453 lets you choose between 5 different predefined scale tunings, or you can create your own custom scale tuning. The PSR-E353 doesn't have those functions-- although it responds to MIDI tuning messages.
And finally, the PSR-E453 has USB audio input/output, which is new to the PSR-E4xx line. The PSR-E353 doesn't.
